Vibrating CPU heatsink / fan assembly

[tomsresults]

Hi,

I recently changed from an Antec900 case to a Thermaltake Element G.

My CPU heatsink is an old Thermaltake Big Typhoon. It's a massive cube of aluminum fins with a 120mm fan attached to it, and its vibrating a lot. The fan shoots air into the CPU and out the sides of the heatsink.

I think the vibrations are travelling and causing the top 200mm case fan to also vibrate up and down, causing a muffled clicking noise a dozen times a second. That fan is also only like an inch away from the giant CPU heat sink and it's not moving much air upward.

If I put my finger on the CPU heatsink and apply pressure the noises stop but if I let go the noises come back in a minute or so.
I tried tightening the heatsink to the motherboard from both sides, but it didn't help much. I think it's just too big and will vibrate no matter what.
I can try making it *really* tight; I guess the mobo and CPU are a few years old so I don't care if I crush them.

Any suggestions? Should I just put in some twist ties to keep the heatsink from moving?

 

[Ubrales]

The system has reached a harmonic frequency (natural frequency) and the vibrations will continue unless this harmonic frequency is changed. The frequency can be changed by adding some stick-on lead weights, like the ones that are used on automotive wheel rims during balancing. Stick some weights at strategic locations until the vibrations stop.

Tightening the fasteners securing the heatsink will not solve this problem.

These lead stick-on weights can also be used on the inside of computer cases to stop the sides from vibrating or rattling.

I did this to my home heating furnace side panel to alter the frequency and thereby stop the harmonic vibrations caused by the blower motor.

This video http://www.youtube.com/watch?v=j-zczJXSxnw describes how natural frequency (harmonic) destroyed the Tacoma Narrows Bridge.

 

[tomsresults]

I may have found a solution. But it'll take a few days to tell since the problem kind of comes and goes.

First, my mobo is really limited at fan control, so I connected my CPU fan to "CHA_FAN1" header and put my rear fan on the "CPU_FAN" header. This enabled me to control the CPU fan.

Next, I ran Speedfan and slowed the CPU fan down.
At 100% power modulation, it is 1395 RPM and causes the annoying vibrating noise.
At 75%, it is 1133 RPM and is silent.
At 55%, it is 870 RPM and causes a very loud buzzing! The buzzing doesn't happen at 50% or 60%.

So I went with 75%.

 

[tomsresults]

So is this problem really resonance? Caused by mechanical vibration?

Or resonance from airflows interacting?

Is it a bad ball bearing?

 

[Ubrales]

So is this problem really resonance? Caused by mechanical vibration?

Or resonance from airflows interacting?

Is it a bad ball bearing?

It is resonance - mechanical resonance. Not air flow related. At the right RPM, the vibrations will stop. The resonant frequency can be controlled by either altering the speed (RPM) or by altering the mass (weights), or by doing both.

Air will create noise only at Mach speeds (air exiting the barrel of a rifle, aircraft breaking the sound barrier) - not a problem in heatsink air flow.

And yes, definitely connect the heatsink fan to the "CPU_FAN" header. This way, the BIOS will control the fan speed depending on the temps, and also, there will be protection if temps rise excessively high (in the case of a fan failure).